
Are you a QOTWer? Do you want to start a thread that isn't a direct answer to the current QOTW? Then this place, gentle poster, is your friend.
( , Sun 1 Apr 2001, 1:00)
« Go Back | Popular

Active/Active Failover = Good! Pretty.
Active/Passive Failover = Meh...
Active/Passive that doesn't fucking work = GRRRR.
INCOMPETENCE!
( , Thu 24 Sep 2009, 15:26, 14 replies, latest was 16 years ago)

We have servers running here, one set of "Active" servers, i.e the ones processing the requests. We also have a set of "Passive" servers - the ones who will handle the requests if the Active ones go down.
Active/Passive works, but you've got a lead-time for it to detect one went down and remove it from the pool. However, it doesn't bloody work as I found out today when the whole bloody lot went down.
This is why we should use active/active.
Sigh.
( , Thu 24 Sep 2009, 15:52, Reply)

Blah blah blah blah blah blah blah go down. Blah blah blah blah blah blah blah went down.
( , Thu 24 Sep 2009, 15:56, Reply)

Active/Active Failover = Good! Pretty.
Active/Passive Failover = Meh...
Active/Passive that doesn't fucking work = GRRRR.
= GIBBERISH
( , Thu 24 Sep 2009, 15:57, Reply)

A resource cluster that does not fail over...
...is not a cluster.
/fellow sysadmin geek
( , Fri 25 Sep 2009, 5:38, Reply)

It is so true... Here they were attempting to use RedHat ClusterFuck, much to my disdain. I'm going to have stern words with them on Monday about doing. it. right.
I cant understand why, when we have such a high load application they are wasting 50% of the resources on standby failover. They're not even running it stepped back, it's just completely passive.
Did i mention the app they created is so monstrous it takes FOUR seconds to respond!?
The worst part is I'm 20, 10 years younger than them.
( , Fri 25 Sep 2009, 10:35, Reply)
« Go Back | Reply To This »